Parameter Constrained Neural Network

Model overview

Parameter constrained neural network classifier using PyTorch with configurable hidden layers to predict taxonomy IDs from product tags. Each taxonomy path is encoded as a p-adic integer and the model minimizes p-adic distance.

5 CV folds

11.28% Mean accuracy

0.0887 Mean F1

0.59801551 Mean p-adic loss

Total train samples: 92,388
Total test samples: 23,097

Key hyperparameters

Prime base(s): 83
Hidden layer sizes: 27
Max tags used: 32

Cross-validation results

Fold	Accuracy	F1	P-adic loss (mean)	Details
0	11.59%	0.0889	0.598354	View details →
1	10.69%	0.0837	0.591221	View details →
2	11.22%	0.0911	0.611152	View details →
3	11.88%	0.0920	0.587771	View details →
4	11.04%	0.0877	0.601580	View details →